On Sat, 25 Sep 2010 00:46:03 +0530
"Aneesh Kumar K. V" [off-list ref] wrote:

On Fri, 24 Sep 2010 11:54:23 -0400, Jeff Layton [off-list ref] wrote:

quoted

On Fri, 24 Sep 2010 18:18:11 +0530
"Aneesh Kumar K.V" [off-list ref] wrote:

quoted

From: Andreas Gruenbacher <redacted>

Some permission models distinguish between the permission to create a
non-directory and a directory.  Pass this information down to
inode_permission() as mask flags

Signed-off-by: Andreas Gruenbacher <redacted>
Signed-off-by: Aneesh Kumar K.V <redacted>
---
 fs/namei.c         |   21 ++++++++++++---------
 include/linux/fs.h |    2 ++
 2 files changed, 14 insertions(+), 9 deletions(-)

diff --git a/fs/namei.c b/fs/namei.c
index b0b8a71..ed786b2 100644
--- a/fs/namei.c
+++ b/fs/namei.c

@@ -253,7 +253,8 @@ int generic_permission(struct inode *inode, int mask,
  * for filesystem access without changing the "normal" uids which
  * are used for other things.
  *
- * When checking for MAY_APPEND, MAY_WRITE must also be set in @mask.
+ * When checking for MAY_APPEND, MAY_CREATE_FILE, MAY_CREATE_DIR,
+ * MAY_WRITE must also be set in @mask.
  */
 int inode_permission(struct inode *inode, int mask)
 {

@@ -1337,13 +1338,15 @@ static int may_delete(struct inode *dir,struct dentry *victim,int isdir)
  *  3. We should have write and exec permissions on dir
  *  4. We can't do it if dir is immutable (done in permission())
  */
-static inline int may_create(struct inode *dir, struct dentry *child)
+static inline int may_create(struct inode *dir, struct dentry *child, int isdir)

									^^^^^
								nit: maybe saner as a bool?

quoted

 {
+	int mask = isdir ? MAY_CREATE_DIR : MAY_CREATE_FILE;
+
 	if (child->d_inode)
 		return -EEXIST;
 	if (IS_DEADDIR(dir))
 		return -ENOENT;
-	return inode_permission(dir, MAY_WRITE | MAY_EXEC);
+	return inode_permission(dir, MAY_WRITE | MAY_EXEC | mask);
 }
 
 /*

@@ -1391,7 +1394,7 @@ void unlock_rename(struct dentry *p1, struct dentry *p2)
 int vfs_create(struct inode *dir, struct dentry *dentry, int mode,
 		struct nameidata *nd)
 {
-	int error = may_create(dir, dentry);
+	int error = may_create(dir, dentry, 0);
 
 	if (error)
 		return error;

@@ -1953,7 +1956,7 @@ EXPORT_SYMBOL_GPL(lookup_create);
 
 int vfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t dev)
 {
-	int error = may_create(dir, dentry);
+	int error = may_create(dir, dentry, 0);
 
 	if (error)
 		return error;

@@ -2057,7 +2060,7 @@ SYSCALL_DEFINE3(mknod, const char __user *, filename, int, mode, unsigned, dev)
 
 int vfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
 {
-	int error = may_create(dir, dentry);
+	int error = may_create(dir, dentry, 1);
 
 	if (error)
 		return error;

@@ -2342,7 +2345,7 @@ SYSCALL_DEFINE1(unlink, const char __user *, pathname)
 
 int vfs_symlink(struct inode *dir, struct dentry *dentry, const char *oldname)
 {
-	int error = may_create(dir, dentry);
+	int error = may_create(dir, dentry, 0);
 
 	if (error)
 		return error;

@@ -2415,7 +2418,7 @@ int vfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *new_de
 	if (!inode)
 		return -ENOENT;
 
-	error = may_create(dir, new_dentry);
+	error = may_create(dir, new_dentry, S_ISDIR(inode->i_mode));

						^^^^ this is a little
						scary, but even if it's
						a directory, it'll get
						kicked out in a later
						check. Would it be
						clearer to move up the
						S_ISDIR() check in this
						function and then pass
						this in as false?

Can you elaborate on this ? 

-aneesh

Hardlinked directories are a no-no, of course. So when I first saw this
patch, it gave me pause. There's a later check in vfs_link though that
explicitly rejects hardlinking directories, so the above is harmless.
It may be more efficient to go ahead and return error if the target
is a directory however and bypass the permission check.

OTOH, maybe there's good reason to do it this way or I'm just being
excessively nitpicky.

-- 
Jeff Layton [off-list ref]